Dan Goldhaber
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Dan Goldhaber has a total of 116 co-authors. Here are the top co-authors by number of shared publications:
- Roddy Theobald
American Institutes for Research, Seattle, WA, USA
19 shared publications - John M. Krieg
Western Washington University
9 shared publications - James Cowan
American Institutes for Research
8 shared publications - Elise Dizon-Ross
CALDER, American Institutes for Research
6 shared publications - Grace T. Falken
Brown University
6 shared publications - Michael DeArmond6 shared publications
Latest Publications
- 1 October 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Can Recognition Improve Results? A First Look at the Massachusetts Student Achievement Award Program
Students from disadvantaged backgrounds are underrepresented among top performers in mathematics and English language arts (ELA) tests, and those who do become top performers are more likely than similarly high-achieving peers to lose ground over time. This study examines the Massachusetts Student Achievement Awards (MSAA) program, which recognized disadvantaged students who had scores judged to be “exceeding expectations” on the Massachusetts state assessment—roughly the top 7% on state standardized tests in mathematics or ELA—with certificates, letters, informational resources, and a modest gift card. Using a difference-in-differences design, we estimate the program's effects on high-achieving, disadvantaged students in Grades 3–7. Recognition increased next-year mathematics test achievement by 0.05–0.07 standard deviations, a result that holds consistently across specifications; effects on ELA achievement are smaller and not consistently significant across model specifications. Within mathematics, effects are larger for girls and for students who were not consistently among the state's top performers. These findings suggest that even among a small population of exceptionally high-performing but under-recognized disadvantaged students, a simple, low-cost, and readily scalable recognition intervention can meaningfully improve short-term academic outcomes—particularly in mathematics.
- 1 October 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Can Recognition Improve Results? A First Look at the Massachusetts Student Achievement Award Program
Students from disadvantaged backgrounds are underrepresented among top performers in mathematics and English language arts (ELA) tests, and those who do become top performers are more likely than similarly high-achieving peers to lose ground over time. This study examines the Massachusetts Student Achievement Awards (MSAA) program, which recognized disadvantaged students who had scores judged to be “exceeding expectations” on the Massachusetts state assessment—roughly the top 7% on state standardized tests in mathematics or ELA—with certificates, letters, informational resources, and a modest gift card. Using a difference-in-differences design, we estimate the program's effects on high-achieving, disadvantaged students in Grades 3–7. Recognition increased next-year mathematics test achievement by 0.05–0.07 standard deviations, a result that holds consistently across specifications; effects on ELA achievement are smaller and not consistently significant across model specifications. Within mathematics, effects are larger for girls and for students who were not consistently among the state's top performers. These findings suggest that even among a small population of exceptionally high-performing but under-recognized disadvantaged students, a simple, low-cost, and readily scalable recognition intervention can meaningfully improve short-term academic outcomes—particularly in mathematics.
- 1 September 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
New Evidence from the Early Teacher Pipeline
Few school-based factors matter more to student success than teacher quality.¹ Because teachers are so important, the policies that shape the teacher workforce—teacher licensure, for example—are critical issues in education policy and research. But before prospective teachers can be licensed, they first need to choose to pursue teaching and complete an approved teacher education program. Largely because of data limitations, these early stages of the teacher pipeline have received relatively less policy and research attention. This is a problem because the decisions people make before entering teacher preparation programs ultimately help determine who is available to join the teacher workforce. To help learn more about the early teacher pipeline, a unique partnership between five university-based teacher education programs, the state of Washington, and researchers gathered a rich array of evidence on teacher preparation programs and their implications for the teacher workforce. This CALDER research brief summarizes the partnership’s findings in three sections: Section 1 examines who enters teacher education programs and how this shapes the pool of potential teachers. Section 2 examines the information teacher education programs collect about applicants, and the applicant attributes faculty consider most important, to better understand how teacher education programs make admissions decisions. Section 3 examines whether the academic preparation students receive in college predicts who eventually becomes a teacher and how effective they are at raising student test scores.
- 1 September 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
New Evidence from the Early Teacher Pipeline
Few school-based factors matter more to student success than teacher quality.¹ Because teachers are so important, the policies that shape the teacher workforce—teacher licensure, for example—are critical issues in education policy and research. But before prospective teachers can be licensed, they first need to choose to pursue teaching and complete an approved teacher education program. Largely because of data limitations, these early stages of the teacher pipeline have received relatively less policy and research attention. This is a problem because the decisions people make before entering teacher preparation programs ultimately help determine who is available to join the teacher workforce. To help learn more about the early teacher pipeline, a unique partnership between five university-based teacher education programs, the state of Washington, and researchers gathered a rich array of evidence on teacher preparation programs and their implications for the teacher workforce. This CALDER research brief summarizes the partnership’s findings in three sections: Section 1 examines who enters teacher education programs and how this shapes the pool of potential teachers. Section 2 examines the information teacher education programs collect about applicants, and the applicant attributes faculty consider most important, to better understand how teacher education programs make admissions decisions. Section 3 examines whether the academic preparation students receive in college predicts who eventually becomes a teacher and how effective they are at raising student test scores.
- 1 September 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Does Better Academic Preparation Predict Teacher Workforce Outcomes?
We link comprehensive student data on university applications and transcripts from three public universities in Washington to teacher workforce entry, effectiveness (measured by value added), and retention. We find that SAT scores are not strongly related to any of these outcomes, but college students with higher university GPAs are more likely to become teachers, and those who enter the workforce with higher GPAs are more effective as measured by math and English language arts (ELA) value-added. These findings are robust to sample selection adjustments and to checks for selection into different college courses, school districts, and out of the teacher workforce. The relationship between GPA and teacher effectiveness is driven largely by grades in math courses designed specifically for teachers. Performance in regular math courses does not predict effectiveness or retention. However, we find that teachers who take at least one advanced math course (400-level or higher) are more effective in both math and ELA than those who do not. Finally, teachers who expressed interest in education at the time of university application are more likely to remain in the workforce once they enter than those who did not. Together, these results suggest that information from college applications and transcripts carries meaningful signals about future teacher effectiveness and retention.
- 1 September 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Does Better Academic Preparation Predict Teacher Workforce Outcomes?
We link comprehensive student data on university applications and transcripts from three public universities in Washington to teacher workforce entry, effectiveness (measured by value added), and retention. We find that SAT scores are not strongly related to any of these outcomes, but college students with higher university GPAs are more likely to become teachers, and those who enter the workforce with higher GPAs are more effective as measured by math and English language arts (ELA) value-added. These findings are robust to sample selection adjustments and to checks for selection into different college courses, school districts, and out of the teacher workforce. The relationship between GPA and teacher effectiveness is driven largely by grades in math courses designed specifically for teachers. Performance in regular math courses does not predict effectiveness or retention. However, we find that teachers who take at least one advanced math course (400-level or higher) are more effective in both math and ELA than those who do not. Finally, teachers who expressed interest in education at the time of university application are more likely to remain in the workforce once they enter than those who did not. Together, these results suggest that information from college applications and transcripts carries meaningful signals about future teacher effectiveness and retention.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Where Have All the Zeros Gone? The Landscape of No Zeros Grading Policies
Debates about grading reform are intensifying, yet little is known about the formal policies that govern how teachers assign grades. This paper examines one prominent example – no-zeros policies (NZPs), which prohibit assigning zeros for assignments or course grades – by analyzing policy documents from school districts in Washington State and linking them to administrative data on district and school characteristics. We find that NZP adoption has more than doubled over the past decade, from roughly 30 districts in 2015-16 to nearly 70 in 2024-25, with the sharpest growth at the high school level. We develop a typology revealing substantial variation in how NZPs are structured, where they appear, and how they are communicated to families. Perhaps most striking, despite being framed as equity-promoting reforms, NZPs are less common in districts serving higher proportions of low-income students. We discuss implications for research and policy and call for greater attention to the design and distribution of grading policies as a foundation for evaluating their impact on students.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Where Have All the Zeros Gone? The Landscape of No Zeros Grading Policies
Debates about grading reform are intensifying, yet little is known about the formal policies that govern how teachers assign grades. This paper examines one prominent example – no-zeros policies (NZPs), which prohibit assigning zeros for assignments or course grades – by analyzing policy documents from school districts in Washington State and linking them to administrative data on district and school characteristics. We find that NZP adoption has more than doubled over the past decade, from roughly 30 districts in 2015-16 to nearly 70 in 2024-25, with the sharpest growth at the high school level. We develop a typology revealing substantial variation in how NZPs are structured, where they appear, and how they are communicated to families. Perhaps most striking, despite being framed as equity-promoting reforms, NZPs are less common in districts serving higher proportions of low-income students. We discuss implications for research and policy and call for greater attention to the design and distribution of grading policies as a foundation for evaluating their impact on students.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
A Broader View of Recovery: Exploring Non-Test Student Outcomes in the Wake of the COVID-19 Pandemic
The COVID-19 pandemic sharply reduced student test scores, and ESSER-funded academic interventions largely failed to recover those losses. Yet few researchers have examined whether these interventions affected students in ways beyond test scores. We use rich longitudinal data from Guilford County Schools (GCS) to document trends before and after the pandemic in attendance, suspensions, disciplinary referrals, and course grades from 2016-17 through 2023-24. All non-test outcomes worsened in GCS after the pandemic, widening pre-existing gaps by race and school poverty level. We also estimate how participation in GCS’s tutoring and summer school in 2021–22 and 2022–23 affect these non-test outcomes. Students who participated in recovery interventions showed improved attendance and discipline in the following year. Effects varied across interventions: summer school robustly improved attendance the following school year, while tutoring improved discipline in one year (2021–22) and attendance in the other (2022–23). These results demonstrate that academic interventions support students on dimensions beyond traditionally measured academic performance and contribute to recovering student engagement and well-being.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
A Broader View of Recovery: Exploring Non-Test Student Outcomes in the Wake of the COVID-19 Pandemic
The COVID-19 pandemic sharply reduced student test scores, and ESSER-funded academic interventions largely failed to recover those losses. Yet few researchers have examined whether these interventions affected students in ways beyond test scores. We use rich longitudinal data from Guilford County Schools (GCS) to document trends before and after the pandemic in attendance, suspensions, disciplinary referrals, and course grades from 2016-17 through 2023-24. All non-test outcomes worsened in GCS after the pandemic, widening pre-existing gaps by race and school poverty level. We also estimate how participation in GCS’s tutoring and summer school in 2021–22 and 2022–23 affect these non-test outcomes. Students who participated in recovery interventions showed improved attendance and discipline in the following year. Effects varied across interventions: summer school robustly improved attendance the following school year, while tutoring improved discipline in one year (2021–22) and attendance in the other (2022–23). These results demonstrate that academic interventions support students on dimensions beyond traditionally measured academic performance and contribute to recovering student engagement and well-being.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Evidence on Team-Based Teaching Models in Arizona
Strategic staffing models are scaling rapidly nationwide. We study the impacts of a team teaching model implemented in one large district in Arizona over several years. We exploit the staggered rollout over time in a difference-in-differences design to estimate effects on student test achievement. We find overall null impacts on student learning, but important heterogeneity by grade and subject. In particular, we find consistently positive effects on middle school math achievement across all specifications, but negative effects at the elementary level in some specifications. Across grades and subjects, teamed teachers also have significantly lower turnover rates and higher evaluation ratings, but the patterns of effects are not consistent with the achievement results. The student achievement findings are broadly consistent with the literature on teacher-level departmentalization and suggest that benefits to team teaching can emerge in contexts where teaming increases collaboration without disrupting established teacher-student relationships, such as in already-departmentalized middle school grades.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Evidence on Team-Based Teaching Models in Arizona
Strategic staffing models are scaling rapidly nationwide. We study the impacts of a team teaching model implemented in one large district in Arizona over several years. We exploit the staggered rollout over time in a difference-in-differences design to estimate effects on student test achievement. We find overall null impacts on student learning, but important heterogeneity by grade and subject. In particular, we find consistently positive effects on middle school math achievement across all specifications, but negative effects at the elementary level in some specifications. Across grades and subjects, teamed teachers also have significantly lower turnover rates and higher evaluation ratings, but the patterns of effects are not consistent with the achievement results. The student achievement findings are broadly consistent with the literature on teacher-level departmentalization and suggest that benefits to team teaching can emerge in contexts where teaming increases collaboration without disrupting established teacher-student relationships, such as in already-departmentalized middle school grades.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Exploring the Landscape of Teacher Applications
Despite widespread concern about teacher shortages, there is little evidence on when teaching positions are posted and how many people apply for them. Using detailed job-posting and application data on nearly 57,000 applicants from 19 traditional school districts and 24 charter organizations, we provide descriptive evidence on the timing of job postings, the supply of applicants, and the characteristics of applicants, emphasizing how each varies across organizations and between the district and charter sectors. We find that late hiring is less prevalent than prior single-district studies suggest—about 11% of openings are posted late—but varies widely across organizations. Applicant supply differs widely across organizations and by sector: districts average 3.3 applicants per opening, versus 6.1 for charters. We also find sectoral differences in applicant pool composition; compared to charters, district applicant pools have higher percentages of White applicants and applicants with prior teaching experience.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
Exploring the Landscape of Teacher Applications
Despite widespread concern about teacher shortages, there is little evidence on when teaching positions are posted and how many people apply for them. Using detailed job-posting and application data on nearly 57,000 applicants from 19 traditional school districts and 24 charter organizations, we provide descriptive evidence on the timing of job postings, the supply of applicants, and the characteristics of applicants, emphasizing how each varies across organizations and between the district and charter sectors. We find that late hiring is less prevalent than prior single-district studies suggest—about 11% of openings are posted late—but varies widely across organizations. Applicant supply differs widely across organizations and by sector: districts average 3.3 applicants per opening, versus 6.1 for charters. We also find sectoral differences in applicant pool composition; compared to charters, district applicant pools have higher percentages of White applicants and applicants with prior teaching experience.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
High School Academic Acceleration and Early Postsecondary Outcomes: Evidence from Washington State
Taking advanced courses in high school predicts positive later outcomes, yet many seemingly prepared students do not take such courses. Acceleration policies that automatically enroll high achieving high school students into more advanced courses have been shown to increase advanced coursetaking, but little is known about the longer-term impacts of these policies. We studied districts in Washington that implemented Academic Acceleration policies, which identified proficient 11th and 12th grade students for automatic enrollment into advanced and dual credit courses. Using a difference-in-differences design, we examined the effects of Academic Acceleration on college enrollment and early postsecondary coursetaking. We found no significant effects of the policy on students’ college enrollment, first-year college coursetaking, or persistence to a second year of college.
- 1 August 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
High School Academic Acceleration and Early Postsecondary Outcomes: Evidence from Washington State
Taking advanced courses in high school predicts positive later outcomes, yet many seemingly prepared students do not take such courses. Acceleration policies that automatically enroll high achieving high school students into more advanced courses have been shown to increase advanced coursetaking, but little is known about the longer-term impacts of these policies. We studied districts in Washington that implemented Academic Acceleration policies, which identified proficient 11th and 12th grade students for automatic enrollment into advanced and dual credit courses. Using a difference-in-differences design, we examined the effects of Academic Acceleration on college enrollment and early postsecondary coursetaking. We found no significant effects of the policy on students’ college enrollment, first-year college coursetaking, or persistence to a second year of college.
- 1 June 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
What Do Educator Job Applications Reveal About Preferences for Schools?
Educator labor markets vary considerably within and across school districts, yet direct evidence of how prospective educators weigh school characteristics when making job decisions is limited. Using applicant tracking data from 19 U.S. school districts, we analyze two distinct forms of applicant choice: initial application decisions to single-school job postings and explicit school preference indicators within multi-school job postings. We find that single-school postings in districts serving higher proportions of historically disadvantaged students receive fewer applications, but that this pattern does not hold across schools within a district. In our analysis of applicants’ choices across schools associated with the same job posting, we find applicants tend to indicate a good deal of flexibility regarding their willingness to work in different schools. However, schools with higher percentages of White students and schools in higher-income neighborhoods were significantly more likely to be marked as preferred. We also find evidence of preference heterogeneity: Asian and Black applicants are significantly more likely than White applicants to indicate a preference for working in schools located in lower-income neighborhoods, serving higher percentages of students eligible for free or reduced-price lunch (FRPL), and higher percentages of Black students.
- 1 June 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
What Do Educator Job Applications Reveal About Preferences for Schools?
Educator labor markets vary considerably within and across school districts, yet direct evidence of how prospective educators weigh school characteristics when making job decisions is limited. Using applicant tracking data from 19 U.S. school districts, we analyze two distinct forms of applicant choice: initial application decisions to single-school job postings and explicit school preference indicators within multi-school job postings. We find that single-school postings in districts serving higher proportions of historically disadvantaged students receive fewer applications, but that this pattern does not hold across schools within a district. In our analysis of applicants’ choices across schools associated with the same job posting, we find applicants tend to indicate a good deal of flexibility regarding their willingness to work in different schools. However, schools with higher percentages of White students and schools in higher-income neighborhoods were significantly more likely to be marked as preferred. We also find evidence of preference heterogeneity: Asian and Black applicants are significantly more likely than White applicants to indicate a preference for working in schools located in lower-income neighborhoods, serving higher percentages of students eligible for free or reduced-price lunch (FRPL), and higher percentages of Black students.
- 1 May 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
A Widening Chasm: The Divergent Paths of High- and Low-Achieving Students in Algebra I After the Pandemic
We follow four cohorts of students in Texas from grades 3 through 9 to identify which factors predict whether students pass the Algebra I end-of-course test by ninth grade. We find that early achievement accounts for most of the variation: racial and socioeconomic gaps in Algebra I completion trace back to differences in third-grade test scores. Because we follow students over time, we can see how the pandemic compounded existing inequalities. First, we find that the pandemic widened racial and socioeconomic gaps within each achievement level—students from disadvantaged groups who started third grade with the same test scores as their more advantaged peers fell further behind than those peers during COVID. Second, the pandemic amplified gaps between achievement levels. The lowest-scoring students in every group experienced steeper declines than their higher-scoring counterparts, with the sharpest losses concentrated among low-achieving students. In short, we find that algebra achievement gaps are rooted in early elementary school and became deeper during the pandemic. Our results underscore the importance of targeting resources to early elementary math instruction, especially for low-achieving students.
- 1 May 2026
- Published by Center for Analysis of Longitudinal Data in Educational Research
A Widening Chasm: The Divergent Paths of High- and Low-Achieving Students in Algebra I After the Pandemic
We follow four cohorts of students in Texas from grades 3 through 9 to identify which factors predict whether students pass the Algebra I end-of-course test by ninth grade. We find that early achievement accounts for most of the variation: racial and socioeconomic gaps in Algebra I completion trace back to differences in third-grade test scores. Because we follow students over time, we can see how the pandemic compounded existing inequalities. First, we find that the pandemic widened racial and socioeconomic gaps within each achievement level—students from disadvantaged groups who started third grade with the same test scores as their more advantaged peers fell further behind than those peers during COVID. Second, the pandemic amplified gaps between achievement levels. The lowest-scoring students in every group experienced steeper declines than their higher-scoring counterparts, with the sharpest losses concentrated among low-achieving students. In short, we find that algebra achievement gaps are rooted in early elementary school and became deeper during the pandemic. Our results underscore the importance of targeting resources to early elementary math instruction, especially for low-achieving students.